Journal of Nuclear Medicine, published on February 23, 2017 ......2017/02/22 · good binding characteristics in rodents, pigs, and a preliminary human study (14,16-19). Here, we

PET Imaging Evaluation of Four Novel Sigma-1 Radiotracers in Nonhuman Primates

Evan Baum1, Zhengxin Cai1, Frederic Bois1, Daniel Holden1, Shu-fei Lin1, Teresa Lara-Jaime1,

Michael Kapinos1, Yuanyuan Chen2, Winnie Deuther-Conrad3, Steffen Fischer3, Sladjana Dukic-

Stefanovic3, Paul Bunse4, Bernhard Wünsch4, Peter Brust3, Hongmei Jia2, Yiyun Huang1*

1PET Center, Department of Radiology and Biomedical Imaging, Yale University School of

Medicine, New Haven, CT, USA

2Ministry of Education Key Laboratory of Radiopharmaceuticals, College of Chemistry, Beijing

Normal University, Beijing, China

3Helmholtz-Zentrum Dresden-Rossendorf, Institute of Radiopharmaceutical Cancer Research,

Leipzig, Germany

4Westfälische Wilhelms-Universität Münster, Department of Pharmaceutical and Medicinal

Chemistry, Münster, Germany

Corresponding author: Yiyun (Henry) Huang, Yale PET Center, 801 Howard Ave, New

Haven, CT 06520-8048. E-mail: [email protected] Telephone: 203-785-3193 Fax: 203-

785-2994

Financial support: This work was supported in part by the National Natural Science Foundation

of China (No. 21471019).

Short running title: Evaluation of Four Sigma-1 PET Tracers

Journal of Nuclear Medicine, published on February 23, 2017 as doi:10.2967/jnumed.116.188052

1

ABSTRACT

Sigma-1 receptors (S1R) are implicated in a variety of diseases including Alzheimer’s disease

and cancer. Previous positron emission tomography (PET) S1R radiotracers are characterized by

slow kinetics or off-target binding that impedes their use for human brain imaging. Here, we

report the first PET imaging evaluation in rhesus monkeys of four 18F-labeled spirocyclic

piperidine-based PET radiotracers (18F-1 to 18F-4) that have previously shown good binding

characteristics in rodents and pigs. Methods: Baseline scans for the four radiotracers were

obtained on an adult male rhesus monkey. Blocking scans were performed with administration of

the S1R selective agonist SA4503 before injection of 18F-2 and 18F-4. Arterial input functions

were measured and binding parameters were determined with kinetic modeling analysis.

Results: In the rhesus brain, all four radiotracers showed high and fast uptake. Tissue activity

washout was rapid for 18F-2 and 18F-4, and much slower for 18F-1 and 18F-3, in line with their

respective in vitro S1R binding affinities. Both the 1-tissue compartment (1TC) and multilinear

analysis-1 (MA1) kinetic models provided good fits of time-activity curves (TACs) and reliable

estimates of distribution volume (VT). Regional VT values were highest in the cingulate cortex

and lowest in the thalamus for all radiotracers. 18F-4 showed greater differential uptake across

brain regions and three-fold higher binding potential (BPND) than 18F-2. SA4503 at the dose of

0.5 mg/kg blocked ~85% (18F-2) and ~95% (18F-4) of radiotracer binding. Conclusion: Tracers

18F-2 and 18F-4 displayed high brain uptake and fast tissue kinetics, with 18F-4 having higher

specific binding signals than 18F-2 in the same monkey. Taken together, these data indicate that

both radiotracers 18F-2 and 18F-4 possess the requisite kinetic and imaging properties as viable

PET tracers for imaging S1R in the primate brain.

Key Words: sigma-1 receptor; PET; radioligand; fluorine-18; rhesus monkey

2

INTRODUCTION

S1R are chaperone proteins localized at mitochondrial associated endoplasmic reticulum

membranes that have been shown to play a role in a wide range of diseases, including addiction,

amnesia, Alzheimer’s disease, amyotrophic lateral sclerosis, and cancer (1,2). Several studies

have detailed the roles of S1R in regulating potassium channels, neuritogenesis, calcium

signaling, memory, and drug addiction (3-5). A variety of psychoactive chemicals and

neurosteroids have been shown to interact with S1R, including haloperidol, (+)-N-

allylnormetazocine (SKF-10,047), cocaine, and progesterone (6-8). Maurice et al. (9)

demonstrated that the S1R agonists (+)-pentazocine, PRE-084, and SA4503 exhibit anti-amnesia

effects in a dose dependent manner in mice with β25-35-amyloid induced amnesia. These results

demonstrate a link between S1R and the pathologic states affecting the cholinergic and

glutamatergic systems, which may be of therapeutic importance in the process of aging (9).

Given the diverse interactions of S1R in pathophysiology, a PET imaging agent for use in

humans would allow for the noninvasive investigation of S1R in vivo and lead to new

understandings of its function and dysfunction in disease states. It will also make it possible to

correlate and translate preclinical findings in animal models to humans and help in the

development of novel therapeutic agents.

Several PET radioligands for S1R have been developed, including 18F-FPS, 18F-FBP,

18F-FTC-146, and 11C-SA4503 (Fig. 1) (10-13). An ideal tracer would possess appropriate

affinity, high selectivity for S1R versus S2R and the vesicular acetylcholine transporter

(VAChT), alongside with fast, reversible tissue kinetics. Furthermore, it should have the

requisite lipophilicity (Log D = 1-3) to cross the blood brain barrier (14). (R)-(+) and (S)-(–)-18F-

fluspidine (18F-1 and 18F-2, respectively) demonstrated good S1R binding affinity (18F-1 Ki =

3

0.57 nM; 18F-2 Ki = 2.30 nM) and favorable kinetics when tested in rodents and pigs (15-17).

Kranz et al. (18) also evaluated 18F-2 in four healthy human subjects for dosimetry calculations.

Li et al. (14) described a series of spirocyclic piperidine derivatives with subnanomolar affinity

for S1R and >100 fold selectivity over S2R and VAChT. In a subsequent paper, Chen et al. (19)

described the synthesis and evaluation of 18F-1’-(4-(2-fluoroethoxy)benzyl)-3H-spiro[2-

benzofuran-1,4’-piperidine] (18F-3) and 18F-1’-((6-(2-fluoroethoxy)pyridin-3-yl)methyl)-3H-

spiro[2-benzofuran-1,4’-piperidine] (18F-4) in mice. In baseline scans and blocking studies with

SA4503, faster clearance and greater specific binding was observed for the N-pyridinyl analog

18F-4 than the N-benzyl analog 18F-3. The spirocyclic piperidine series thus demonstrated

promising properties to image S1R with PET based on their high selectivity towards S1R and

good binding characteristics in rodents, pigs, and a preliminary human study (14,16-19). Here,

we report the first PET imaging evaluation in nonhuman primate of four radioligands (18F-1-4,

Fig. 1) from this series to assess their pharmacokinetic and in vivo binding properties, and to

select the most suitable tracer for advancing to evaluation in humans.

4

MATERIALS AND METHODS

Chemistry

Precursors and reference standards for 18F-1 and 18F-2 were prepared at Westfalische

Wilhelms-Universitat Munster, as previously reported (15). Precursors and standards for 18F-3

and 18F-4 were synthesized at Beijing Normal University, as described before (14).

Radiochemistry

Instrumentation for radiochemistry procedures and the production of 18F-fluoride have

been described previously (20,21). Radiosynthesis of 18F-1 to 18F-3 was achieved via

nucleophilic displacement of the corresponding tosylate precursors (5-7, Fig. 2) with 18F- in the

presence of Kyrptofix 2.2.2 and potassium carbonate (14,15). 18F-4 was synthesized in a two-pot,

two-step synthesis, first isolating 18F-fluoroethyl tosylate followed by its reaction with the 2-

pyridinol precursor (8, Fig. 2) (19,22). Chemical purity, radiochemical purity, and specific

activity were determined by high performance liquid chromatography (HPLC) analysis of the

final product solutions. Identities of the labeled compounds were confirmed by co-injection of

the products with their respective unlabeled reference standards.

PET Imaging Experiments in Rhesus Monkeys

PET Procedures. Experiments were performed in rhesus monkeys (Macaca mulatta)

according to procedures approved by the Yale University Institutional Animal Care and Use

Committee and described previously (20).

Three animals were used in this study. The animals were immobilized with ketamine (10

mg/kg intramuscularly) and anesthetized with 1.5-2.5% isoflurane. An arterial line was placed in

the radial or femoral artery for blood sampling. Scans were performed on a FOCUS 220 camera.

Before radioligand injection, a 9-min transmission scan was obtained for attenuation correction.

Baseline scans were obtained over 4 h on a 7-year-old male rhesus monkey (13.8 kg). Each

tracer was injected intravenously over 3 min as a slow bolus (~185 MBq in 10 mL). Two-hour

blocking scans of 18F-2 and 18F-4 were performed with a dose of SA4503 (6) (0.5 mg/kg) given

5

intravenously 10 min prior to radioligand administration. Two additional baseline scans were

obtained for 18F-2 and 18F-4 on a 12-year-old female (6.1 kg) and a 9-year-old female (9.7 kg)

monkey, respectively, for comparison with baseline scans obtained in the 7-year-old male. Eight

PET scans were obtained in total.

Metabolite Analysis and Arterial Input Function Measurement. Procedures for

measurement of the arterial input function, including sample preparation, metabolite analysis,

and data processing have been described previously (20). Arterial samples were collected at

preselected time points and the radioactivity concentrations in the whole blood and plasma were

measured. During the 4 h baseline scans, samples at 3, 8, 15, 30, 60, 90, 120, 180, and 240 min

after injection were processed and analyzed by analytical HPLC using a modified column-

switching system (23) to determine the fraction of unmetabolized tracer over the course of the

scan. A biexponential function was fitted to the measured parent fractions to produce a

continuous function describing the parent fraction over time. The input function was calculated

as the product of the total plasma activity and interpolated parent fraction at each time point. The

measured input function values were fitted to a sum of three exponentials, and the fitted values

were used as inputs for kinetic analyses.

Plasma Free Fraction and Log D Measurement. The free fraction in plasma (fP) was

measured via ultrafiltration of 0.3 mL aliquots of plasma spiked with a small amount (~740 kBq)

of radioligand, repeated in triplicate. The amount of radioactivity in the filter and filtrate was

counted, and fP calculated as the ratio of the concentration (radioactivity/mL) of the filtrate to the

total activity. The Log D of each tracer was determined by the shake-flask method as described

previously (24).

Image Analysis and Kinetic Modeling. Procedures for PET image reconstruction,

definition of regions of interest (ROIs), and kinetic analysis have been detailed previously (20).

Emission data were attenuation corrected using the transmission scan, and dynamic images (33

frames over 120 min or 57 frames over 240 min) were reconstructed using a filtered back-

6

projected algorithm with a Shepp-Logan filter. ROIs were defined from a single representative

anatomic rhesus MR image registered to a template image. Registration parameters were derived

to apply ROIs to each PET scan, and time activity curves (TACs) were generated for the

following 16 cortical and subcortical brain regions: amygdala, brainstem, caudate, cerebellum,

cingulate cortex, frontal cortex, globus pallidus, hippocampus, insula, nucleus accumbens,

occipital cortex, pons, putamen, substantia nigra, temporal cortex, and thalamus.

Regional volumes of distribution (VT, mL∙cm-3) were determined by kinetic analysis of

the TACs, using the metabolite-corrected arterial plasma input function according to 1- and 2-

tissue compartment (1TC, 2TC) models, and the multilinear analysis-1 (MA1) method as

described previously (25,26). Standard errors were compared to determine the optimal model for

regional VT estimates.

Comparison of VT between tracers was used to determine relative regional BPND by

graphical methods (27). In this analysis, of one tracer is plotted on the x-axis and of

another is plotted on the y-axis. A linear regression yields the following equation:

1 (1)

More negative values for the y-intercept indicate > , and vice versa. Since fP was

measured for all tracers, the ratio of equilibrium dissociation constants (KD) can be determined

from the slope of the regression (27).

Sigma-1 receptor occupancies with SA4503 were calculated using VT values from all 16

ROIs to create occupancy plots according to the method of Cunningham et al. (28). For 18F-2

and 18F-4, regional BPND values were calculated using the nondisplaceable volume of

distribution (VND) obtained from the occupancy studies, where BPND = (VT/VND) – 1 (26).

RESULTS

Radiochemistry

7

All tracers were synthesized in > 96% radiochemical purity and high specific activity

(296.9 GBq/µmol average at end of synthesis, n = 7). Total synthesis time was 110 ± 30 min.

Radiochemical yields (±SD where applicable) are 2.0%, 6.2 ± 1.7% (n = 3), 7.8%, and 6.5 ±

9.2% (n = 3), respectively, for 18F-1, 18F-2, 18F-3, and 18F-4.

In Vivo Evaluation in Rhesus Monkeys

After a bolus injection of the tracers (181.5 ± 10.6 MBq; specific activity of 242.4

GBq/µmol average at time of injection; injected mass of 0.45 ± 0.29 µg, n = 7) into an adult male

rhesus monkey, total plasma activity and parent activity exhibited a rapid rise and clearance,

followed by a stabilization or slow decrease over time for 18F-1 to 18F-3, and a slight increase for

18F-4 (Fig. 3A). Metabolism rates were moderate, with 37%, 35%, 18%, and 19% of parent

fraction, respectively, for 18F-1 to 18F-4 at 60 min post injection (Fig. 3B). Blocking with

SA4503 increased plasma activity particularly for 18F-4 (Fig. 3A) and increased metabolism for

18F-2 and 18F-4 (Fig. 3B). Typical metabolite profiles over time under baseline conditions are

presented in Supplemental Fig. 1-4. Polar metabolites have been previously suggested for these

tracers that should not enter the brain and interfere with PET quantitation (14,15,17).

Plasma fP values were measured at 2%, 2%, 8%, and 17%, respectively, for 18F-1-4,

consistent with their respective measured Log D values of 2.80, 2.80, 2.55, and 2.50. A summary

of in vitro Ki, fP, and Log D values are shown in Table 1.

Regional TACs (Fig. 4A-D) were generated and analyzed with 1TC and 2TC models

(26), as well as the MA1 method (25) using the metabolite corrected arterial plasma input

function. The 1TC model was found to provide better fits than the 2TC model, with the 2TC

model producing high standard error (SE) across many regions (e.g. >20% SE in 28% of regions

under all conditions, and >150% SE in 75% of regions for 18F-4 under blocking condition).

Therefore, the 1TC would be considered an appropriate model for analysis of imaging data.

Regional VT values estimated by MA1 showed good correlation with 1TC values (e.g. for 18F-4,

VT (MA1) = 0.975 VT (1TC) + 1.239, r2 = 0.996). Listed in Table 2 are the 1TC-derived VT values for

8

the tracers across brain regions, including those under blocking conditions. Values under

baseline conditions are generally consistent with regional distribution trends seen in the TACs.

Blocking with 0.5 mg/kg SA4503 for 18F-2 and 18F-4 reduced regional disparities, with VT in

high uptake regions trending down to levels seen in low binding regions (Table 2). The time

stability of VT values was also determined for the four tracers, revealing bias and regional errors

associated with shorter scan times (Fig. 5).

Administration of the S1R agonist SA4503 blocked 85% and 95% of specific binding of

18F-2 and 18F-4, respectively (Supplemental Fig. 5). Values of VND were calculated for 18F-2 and

18F-4 from the occupancy plots using 1TC and MA1 derived VT values (28). This analysis

yielded VND (1TC) = 6.87 and VND (MA1) = 7.30 for 18F-2, and VND (1TC) = 9.29 and VND (MA1) =

10.44 for 18F-4. These values were then used to calculate the binding potential (BPND), as a

measure of specific binding signal, across the brain regions (Table 3). Relative BPND and KD for

all tracers was also assessed by the graphical methods of Guo et al. (27) by comparing baseline

VT values, which showed for rank order of BPND: 18F-3 > 18F-1 > 18F-4 > 18F-2 and for KD: 18F-4

> 18F-2 > 18F-1 > 18F-3.

DISCUSSION

In this paper, we describe the evaluation of four 18F-labeled sigma-1 receptor tracers in

nonhuman primates. (R)-(+)- and (S)-(–)-18F-Fluspidine (18F-1 and 18F-2, respectively) were

previously found to have high affinity for S1R, metabolic stability, and appropriate binding

kinetics in mice (16) and pigs (17). 18F-2 was also evaluated in healthy human subjects and

found to have an effective dose (21.0 Sv/MBq) within acceptable imaging limits (18).

Spirocyclic piperidines of related structure (18F-3 and 18F-4) also showed good affinity and

promising binding kinetics for S1R in mice and rats (14,19). Hence, we evaluated these four PET

tracers in nonhuman primates to compare their pharmacokinetic and binding characteristics and

assess feasibility for use in human subjects.

9

In the rhesus monkey brain, all four tracers demonstrated high uptake as shown in TACs

presented in Fig. 4A-D. Activity peaked between 10-20 min post injection, indicating fast uptake

kinetics. Fast tracer washout was observed for 18F-2 and 18F-4, with slow washout for 18F-1 and

18F-3. TACs depict 18F-4 as having the most heterogeneous uptake across brain regions, further

displayed in summed images of the adult male monkey (Fig. 6A-D). For all tracers, regional

TACs show the highest uptake in the cingulate cortex, insula, and frontal cortex; lower in the

hippocampus, temporal and occipital cortices; and lowest in the caudate and thalamus (Fig. 4A-

D). While this distribution pattern is in agreement with results from in vivo imaging study with

11C-SA4503 and in vitro autoradiography study with (+)-3H-3-PPP in rhesus monkey brain

(29,30), it differs somewhat from ex vivo autoradiography studies with 18F-3 and 18F-4 in rats

(14,19), which showed high accumulation in the temporal cortex, frontal cortex, and vermian

lobule of the cerebellum; moderate uptake in the hippocampus, hypothalamus, and thalamus; and

low accumulation in the nucleus accumbens. Studies with 18F-fluspidine in mice (16) showed

highest uptake in the facial nucleus, moderate uptake in the cerebellum, and low binding in the

thalamus and caudate/putamen. PET imaging studies using 18F-1 and 18F-2 in pigs (17)

demonstrated narrow regional differences, with highest uptake in the midbrain, pons, and

thalamus; moderate uptake in the hippocampus, temporal and occipital cortices; and lowest in the

frontal cortex. The distribution patterns of these four new tracers in monkeys also align well with

that of 11C-SA4503 in humans (31), with high uptake in cortical and limbic areas and lower

uptake in the caudate, putamen, and thalamus. Species differences in tracer uptake between

rodents, pigs and primates highlight the importance of tracer evaluation in non-human primates

before translation to humans.

Regional VT values appear to be sex independent as comparison baseline scan for 18F-4 in

a 9-year old female rhesus monkey gave numbers similar to those seen in the 7-year old male

(Table 2). However, the baseline VT values obtained with 18F-2 in the 12-year old female were

higher than those from the 7-year old male, which could be due to individual animal variation in

10

S1R expression, or age effect, as Matsuno et al. reported increased S1R density in aged rhesus

monkeys (29).

A comparison of VT estimates versus scan time demonstrated the bias and regional error

associated with shorter scan times (Fig. 5). With 60 min of scan data, VT values for both 18F-2

and 18F-4 were within 10% difference of those derived from the full 240 min data, and within 5%

difference when estimated with 90 min of scan data. Tracers 18F-1 and 3 demonstrated much

larger bias and greater errors with shorter scan times, only approaching within 5% difference

from the 240 min VT values at 210 min of scanning time. VT values for 18F-1 were

underestimated while those of 18F-3 were overestimated with shorter acquisition times (Fig. 5).

For these reasons, 18F-2 and 18F-4 gave reliable VT estimates at 90 min, while 18F-1 and 18F-3

required much longer scan times.

Based on their extremely slow tissue kinetics, 18F-1 and 18F-3 were deemed unsuitable for

PET neuroimaging of S1R in humans, and therefore not selected for the blocking studies.

Nondisplaceable volume of distribution (VND) for 18F-2 and 18F-4 was estimated from the

occupancy plots and used to calculate BPND, with 1TC-derived values displayed in Table 3.

These values demonstrate higher specific binding signals for 18F-4 than 18F-2. A graphical

comparison of VT as a relative measure of BPND (27) revealed BPND rank order of 18F-3 > 18F-1 >

18F-4 > 18F-2, further supporting the results obtained from the blocking studies. The higher BPND

values for 18F-1 and 18F-3 compared to 18F-2 and 18F-4 are likely due to their lower Ki values

(i.e., higher S1R binding affinity), but higher affinity also contributes to slow and unfavorable

binding kinetics in this instance. A comparison of the relative dissociation constants at

equilibrium yielded KD rank order of 18F-4 > 18F-2 > 18F-1 > 18F-3, which in general are

consistent with in vitro Ki measurements (i.e, Ki values for 18F-4 and 18F-2 are higher than those

of 18F-1 and 18F-3), but also reveals some differences between in vivo KD and in vitro Ki values

(Table 1). In addition to individual and species specific variation, these differences between in

vitro and in vivo affinities could be due to inter-laboratory variations in methodologies and

11

techniques, or temperature effects, as in vitro Ki measurements were performed at room

temperature, while in vivo measurements were at body temperature (37 °C). It should be noted,

however, that this method of graphical comparison is less useful for 18F-1 and 18F-3 because they

display nearly irreversible kinetics.

The irreversible nature of binding for 18F-1 was confirmed in an in vitro experiment to

measure the individual rate constants (kon, koff) and the dissociation constants (KD) for 18F-1 and

18F-2, which generated KD value of 0.099 nM for 18F-2 on cloned human S1R, and kon and koff

values of 3.46 x 10 M-1min-1 and 0.0342 min-1, respectively. However, for 18F-1, koff was

extremely slow and could not be reliably measured (See Supplemental Fig. 6 and 7 and

Supplemental Table 1).

18F-2 and 18F-4 may offer advantages over previously developed S1R PET tracers. 11C-

SA4503 has been studied in rhesus monkeys (12,29), showing similarly high uptake and regional

distributions as 18F-2 and 18F-4, and with reasonable subtype selectivity for S1R (S2R/S1R =

103). 11C-SA4503, however, demonstrates slow washout for the time scale of a 11C-labelled PET

tracer (32) and requires an on-site cyclotron for production. Subsequent studies (33-35) with 11C-

SA4503 showed lower subtype selectivity (S2R/S1R = 13.3-55.0) than previously reported, and

when combined with its affinity for VAChT (Ki = 50 nM, VAChT/S1R = 11.3) (34), it may

exhibit greater nonspecific binding than 18F-2 and 18F-4. A more recent S1R tracer, 18F-FTC-146,

has been evaluated in mice, rats, and squirrel monkeys (36,37). It demonstrated high binding

specificity and selectivity (Ki of 0.0025 nM, 364 nM, and 463 nM, respectively, for S1R, S2R,

and VAChT), as well as favorable kinetics and promising imaging properties in the squirrel

monkey brain (36). Nonetheless, 18F-FTC-146 has been shown to accumulate in the skull of

squirrel monkeys, potentially confounding PET quantitation in the brain and limiting its utility

(36). Defluorination and bone uptake of many 18F-labeled tracers may occur in lower species but

not in humans (38), so further studies are warranted for 18F-FTC-146. Defluorination was not

observed for 18F-2 and 18F-4. Alongside these possible advantages over previously reported S1R

12

tracers, 18F-2 and 18F-4 show high regional BPND values, fast kinetics, and the requisite PET

imaging characteristics to image and quantitate S1R in the primate brain.

CONCLUSIONS

In this report, we compare the binding and kinetic properties of four 18F-labeled

spirocyclic piperidine derivatives in nonhuman primates. Among these tracers, 18F-2 and 18F-4

exhibit favorable metabolic profiles, fast brain uptake kinetics, and high specific binding signals

in rhesus monkeys. Tracer 18F-4 has ten-fold higher fP, three-fold higher BPND, and greater VT

values than 18F-2 when compared in the same monkey. Both tracers also give reliable estimates

of VT with short (90 min) scan times. Taken together, these data indicate that tracers 18F-2 and

18F-4 possess the requisite kinetic and imaging properties as viable PET tracers for imaging S1R

in the primate brain, and thus warrant further evaluation in humans.

DISCLOSURE

None

ACKNOWLEDGMENT

The authors thank the staff at the Yale PET Center for their expert technical assistance.

13

REFERENCES

1. Crottes D, Guizouarn H, Martin P, Borgese F, Soriani O. The sigma-1 receptor: a

regulator of cancer cell electrical plasticity? Front Physiol. 2013;4:175.

2. Maurice T, Su T-P. The pharmacology of sigma-1 receptors. Pharmacol Ther.

2009;124:195-206.

3. Hayashi T, Su T-P. Sigma-1 receptors at galactosylceramide-enriched lipid

microdomains regulate oligodendrocyte differentiation. Proc Natl Acad Sci U S A.

2004;101:14949-14954.

4. Matsumoto RR, Liu Y, Lerner M, Howard EW, Brackett DJ. Sigma receptors: potential

medications development target for anti-cocaine agents. Eur J Pharmacol. 2003;469:1-12.

5. Maurice T, Lockhart BP. Neuroprotective and anti-amnesic potentials of sigma ()

receptor ligands. Prog Neuropsychopharmacol Biol Psychiatry. 1997;21:69-102.

6. Matsuno K, Nakazawa M, Okamoto K, Kawashima Y, Mita S. Binding properties of

SA4503, a novel and selective σ1 receptor agonist. Eur J Pharmacol. 1996;306:271-279.

7. Tam SW, Cook L. Sigma opiates and certain antipsychotic drugs mutually inhibit (+)-

[3H]SKF 10,047 and [3H]haloperidol binding in guinea pig brain membranes. Proc Natl Acad Sci

U S A. 1984;81:5618-5621.

8. Urani A, Privat A, Maurice T. The modulation by neurosteroids of the scopolamine-

induced learning impairment in mice involves an interaction with sigma1 (1) receptors. Brain

Res. 1998;799:64-77.

9. Maurice T, Su TP, Privat A. Sigma1 (1) receptor agonists and neurosteroids attenuate

B25-35-amyloid peptide-induced amnesia in mice through a common mechanism. Neuroscience.

1998;83:413-428.

14

10. Collier TL, O'Brien JC, Waterhouse RN. Synthesis of [18F]-1-(3-fluoropropyl)-4-(4-

cyanohenoxymethyl)-piperidine: a potential sigma-1 receptor radioligand for PET. J Labelled

Comp Radiopharm. 1996;38:785-794.

11. Shiue CY, Shiue GG, Zhang SX, et al. N-(N-benzylpiperidin-4-yl)-2-

[18F]fluorobenzamide: a potential ligand for PET imaging of sigma receptors. Nucl Med Biol.

1997;24:671-676.

12. Kawamura K, Ishiwata K, Tajima H, et al. In vivo evaluation of [11C]SA4503 as a PET

ligand for mapping CNS sigma1 receptors. Nucl Med Biol. 2000;27:255-261.

13. James ML, Shen B, Zavaleta CL, et al. New positron emission tomography (PET)

radioligand for imaging -1 receptors in living subjects. J Med Chem. 2012;55:8272-8282.

14. Li Y, Wang X, Zhang J, et al. Synthesis and evaluation of novel 18F-labeled spirocyclic

piperidine derivatives as 1 receptor ligands for positron emission tomography imaging. J Med

Chem. 2013;56:3478-3491.

15. Holl K, Falck E, Kohler J, et al. Synthesis, characterization, and metabolism studies of

fluspidine enantiomers. ChemMedChem. 2013;8:2047-2056.

16. Fischer S, Wiese C, Maestrup EG, et al. Molecular imaging of sigma receptors: synthesis

and evaluation of the potent sigma1 selective radioligand [18F]fluspidine. Eur J Nucl Med Mol

Imaging. 2011;38:540-551.

17. Brust P, Deuther-Conrad W, Becker G, et al. Distinctive in vivo kinetics of the new

sigma1 receptor ligands (R)-(+)- and (S)-(-)-18F-fluspidine in porcine brain. J Nucl Med.

2014;55:1730-1736.

15

18. Kranz M, Sattler B, Wüst N, et al. Evaluation of the enantiomer specific biokinetics and

radiation doses of [18F]fluspidine—a new tracer in clinical translation for imaging of σ1

receptors. Molecules. 2016;21:1164.

19. Chen Y-Y, Wang X, Zhang J-M, et al. Synthesis and evaluation of a 18F-labeled

spirocyclic piperidine derivative as promising σ1 receptor imaging agent. Bioorg Med Chem.

2014;22:5270-5278.

20. Zheng MQ, Nabulsi N, Kim SJ, et al. Synthesis and evaluation of 11C-LY2795050 as a

kappa-opioid receptor antagonist radiotracer for PET imaging. J Nucl Med. 2013;54:455-463.

21. Bois F, Gallezot JD, Zheng MQ, et al. Evaluation of [18F]-(-)-

norchlorofluorohomoepibatidine ([18F]-(-)-NCFHEB) as a PET radioligand to image the

nicotinic acetylcholine receptors in non-human primates. Nucl Med Biol. 2015;42:570-577.

22. Schoultz BW, Hjornevik T, Reed BJ, et al. Synthesis and evaluation of three structurally

related 18F-labeled orvinols of different intrinsic activities: 6-O-[18F]fluoroethyl-diprenorphine

([18F]FDPN), 6-O-[18F]fluoroethyl-buprenorphine ([18F]FBPN), and 6-O-[18F]fluoroethyl-

phenethyl-orvinol ([18F]FPEO). J Med Chem. 2014;57:5464-5469.

23. Hilton J, Yokoi F, Dannals RF, Ravert HT, Szabo Z, Wong DF. Column-switching

HPLC for the analysis of plasma in PET imaging studies. Nucl Med Biol. 2000;27:627-630.

24. Wilson AA, Jin L, Garcia A, DaSilva JN, Houle S. An admonition when measuring the

lipophilicity of radiotracers using counting techniques. Appl Radiat Isot. 2001;54:203-208.

25. Ichise M, Toyama H, Innis RB, Carson RE. Strategies to improve neuroreceptor

parameter estimation by linear regression analysis. J Cereb Blood Flow Metab. 2002;22:1271-

1281.

16

26. Innis RB, Cunningham VJ, Delforge J, et al. Consensus nomenclature for in vivo imaging

of reversibly binding radioligands. J Cereb Blood Flow Metab. 2007;27:1533-1539.

27. Guo Q, Owen DR, Rabiner EA, Turkheimer FE, Gunn RN. A graphical method to

compare the in vivo binding potential of PET radioligands in the absence of a reference region:

application to [11C]PBR28 and [18F]PBR111 for TSPO imaging. J Cereb Blood Flow Metab.

2014;34:1162-1168.

28. Cunningham VJ, Rabiner EA, Slifstein M, Laruelle M, Gunn RN. Measuring drug

occupancy in the absence of a reference region: the Lassen plot re-visited. J Cereb Blood Flow

Metab. 2010;30:46-50.

29. Kawamura K, Kimura Y, Tsukada H, et al. An increase of sigma receptors in the aged

monkey brain. Neurobiol Aging. 2003;24:745-752.

30. Mash DC, Zabetian CP. Sigma receptors are associated with cortical limbic areas in the

primate brain. Synapse. 1992;12:195-205.

31. Sakata M, Kimura Y, Naganawa M, et al. Mapping of human cerebral sigma1 receptors

using positron emission tomography and [11C]SA4503. Neuroimage. 2007;35:1-8.

32. Sakata M, Kimura Y, Naganawa M, et al. Shortened protocol in practical [11C]SA4503-

PET studies for sigma1 receptor quantification. Ann Nucl Med. 2008;22:143-146.

33. Lever JR, Gustafson JL, Xu R, Allmon RL, Lever SZ. σ1 and σ2 receptor binding affinity

and selectivity of SA4503 and fluoroethyl SA4503. Synapse. 2006;59:350-358.

34. Ishiwata K, Kawamura K, Yajima K, QingGeLeTu, Mori H, Shiba K. Evaluation of (+)-

p-[11C]methylvesamicol for mapping sigma1 receptors: a comparison with [11C]SA4503. Nucl

Med Biol. 2006;33:543-548.

17

35. Hirata M, Mori T, Soga S, Umeda T, Ohmomo Y. Synthesis and in vitro evaluation of

iodinated derivatives of piperazine as a new ligand for sigma receptor imaging by single photon

emission computed tomography. Chem Pharm Bull (Tokyo). 2006;54:470-475.

36. James ML, Shen B, Nielsen CH, et al. Evaluation of σ-1 receptor radioligand 18F-FTC-

146 in rats and squirrel monkeys using PET. J Nucl Med. 2014;55:147-153.

37. Shen B, James ML, Andrews L, et al. Further validation to support clinical translation of

[18F]FTC-146 for imaging sigma-1 receptors. EJNMMI Res. 2015;5:1-10.

38. Pike VW. PET radiotracers: crossing the blood–brain barrier and surviving metabolism.

Trends Pharmacol Sci. 2009;30:431-440.

18

FIGURE 1. Structures of previous sigma-1 tracers 18F-FPS, 18F-FBP, 11C-SA4503, 18F-FTC-

146, and spirocyclic piperidine derivatives 18F-1 to 18F-4, with in vitro Ki values for S1R (10-15).

19

FIGURE 2. Syntheses of the four radiotracers. Reagents and conditions: a. 18F-, Kryptofix 2.2.2,

K2CO3, MeCN, 85 °C, 20 min; b. 18F-, Kryptofix 2.2.2, K2CO3, MeCN, 95 °C, 20 min; c. 18F-,

Kryptofix 2.2.2, K2CO3, MeCN, 80 °C, 10 min; d. Cs2CO3, DMF, 110 °C, 10 min.

20

FIGURE 3. Plasma analysis of the four radiotracers. (A) Total radioactivity in plasma over time.

(B) Time course of parent fraction from the 4 h baseline and 2 h blocking scans with SA4503.

21

FIGURE 4. Time activity curves of 18F-1 (A), 18F-2 (B), 18F-3 (C), and 18F-4 (D) from the

baseline scans. SUV = standardized uptake value.

22

FIGURE 5. Time stability of VT for the four tracers in nonhuman primate brain. Times refer to

mid-times of each 10 min acquisition. Data from 60 to 240 min were analyzed in 30 min

increments, and VT is expressed as percentage of the value derived with the complete data set

(240 min). Each point is an average VT from 16 ROIs. Deviation from 100% of mean value

indicates bias associated with shorter scanning times, whereas standard deviation indicates

regional error associated with shorter scanning times. The bias and error for 18F-3 at 60 min was

too large (>10,000%) to display.

23

FIGURE 6. PET images (coronal, transverse, and sagittal view) summed from 30-45 min of the

baseline scans for 18F-1 (A), 18F-2 (B), 18F-3 (C), and 18F-4 (D). SUV = standardized uptake

value.

24

FIGURE AND TABLE LEGENDS

TABLE 1 Comparison of binding affinity, selectivity, lipophilicity, and plasma free fraction (fp)

Ligand Ki (sigma-1) Ki (sigma-2) Selectivity Log D fp

18F-1 0.57 nM* 1,650 nM* 2,895* 2.8 2%

18F-2 2.3 nM* 897 nM* 390* 2.8 2%

18F-3 0.79 nM† 277 nM† 351† 2.55 8%

18F-4 2.3 nM† 327 nM† 142† 2.5 18%

*Taken from Holl et al. (15)

†Taken from Li et al. (14)

25

TABLE 2 Comparison of 1TC-derived VT values for the four tracers across different brain regions

Condition

VT (mL∙cm-3)

Amygdala Caudate Cerebellum Cingulate

Cortex Frontal Cortex

Hippocampus Occipital Cotrtex

Putamen Temporal

Cortex Thalamus

18F-1 baseline 152.8 180.8 174.9 291.4 215.5 192.8 174.1 199 215.7 127.7 18F-2 baseline 14.6 (31.6)* 13.5 (28.4)* 13.6 (30.2)* 19.6 (37.9)* 17.8 (35.1)* 16.4 (33.5)* 14.6 (30.0)* 14.9 (31.8)* 16.5 (32.6)* 12.2 (30.2)* 18F-3 baseline 574.9 479.3 507.8 1351.9 1099.7 575.8 786.9 530.3 916.8 471 18F-4 baseline 44.4 (42.3)* 36.7 (35.7)* 35.3 (32.7)* 57.9 (56.8)* 46.5 (47.6)* 48.2 (46.5)* 38.2 (31.7)* 38.1 (35.9)* 44.4 (40.6)* 35.3 (38.9)*

18F-2 SA4503 block 6.2 7.1 7.6 8.6 8.3 7.5 8 8.8 8.3 7.2 18F-4 SA4503 block 9.4 10.3 9.9 12 11.9 9.8 10.8 11.9 11.5 9.8

*baseline VT values in two different female rhesus monkeys are noted in parentheses for 18F-2 and 18F-4

26

TABLE 3 1TC-derived BP

ND values of 18F-2 and 18F-4 across different brain regions

Radioligand

BPND

Amygdala Caudate Cerebellum Cingulate

Cortex Frontal Cortex

Hippocampus Occipital Cortex

Putamen Temporal

Cortex Thalamus

18F-2 1.13 0.96 0.98 1.85 1.59 1.38 1.13 1.17 1.40 0.77 18F-4 3.76 2.94 2.78 5.21 3.99 4.16 3.09 3.08 3.76 2.78

1

Evaluation of Four Sigma-1 PET Tracers: Supplemental

Supplemental Figure 1. Time course for metabolism of 18F-1 in plasma, showing gamma HPLC

chromatograms. The peak at ~11 min is the parent compound, and the rest are metabolites.

HPLC conditions: Contents on the capture column were backwashed onto a Phenomenex Luna

C18 (2) column (5 µm, 4.6 x 250 mm) eluting with 60% acetonitrile/40% 0.1 M ammonium

formate, flow rate = 1.20 mL/min.

2






3






4


chromatograms. The peak at ~10.5 min is the parent compound, and the rest are metabolites.




5

Supplemental Figure 5. Receptor occupancy plots for blocking of 18F-2 and 18F-4 binding with

0.5 mg/kg of SA4503 in the same adult rhesus monkey. Plot of changes in 1T derived VT

between baseline and 120 min blocking scans vs. 240 min baseline scans for sixteen brain

regions for 18F-2 (A) and 18F-4 (B). The slope of the regression is the estimated S1R occupancy

by SA4503, and the x-intercept is the estimated nondisplaceable volume of distribution (VND)

from which BPND can be calculated.

y = 0.8545x - 5.8769R² = 0.8322

0

4

8

12

0 5 10 15 20 25

V TBa

selin

e -V

TBl

ock

VT Baseline

A

y = 0.9528x - 9.3351R² = 0.9516

0

10

20

30

40

50

0 10 20 30 40 50 60 70

V TBa

selin

e -V

TBl

ock

VT Baseline

B

6

Association and dissociation studies of 18F-1 and 18F-2

Methods

Kinetic studies with rat and human sigma-1 receptors were performed using membrane

homogenates obtained from the rat cortex (female SPRD rats, 10-12 weeks old) and HEK293

cells stably transfected with human SIGMAR1 (by courtesy of Olivier Soriani, Institute of

Biology Valrose, UNS Université Nice Sophia Antipolis, France), respectively, as well as

identical batch of the respective radioligand 18F-1 [(R)-18F-fluspidine] or 18F-2 [(S)-18F-

fluspidine]. Association and dissociation experiments were conducted at room temperature in

incubation buffer (50 mM TRIS, pH 7.4, 120 mM NaCl, 5 mM KCl, 2 mM CaCl2, 1 mM

MgCl2). Non-specific binding was determined by addition of 1 µM haloperidol in the incubation

buffer.

Association studies were started with the application of radioligand, and receptor-bound

ligand separated from free ligand by filtration (GF-B glass-fibre filter; 48-sample harvester,

Brandel, Gaithersburg, MD, USA) after 0.5, 1, 3, 5, 7, 10, 15, 20, 30, 45, 60, 90, 120, and 180

min incubation.

For dissociation studies, receptor preparation and 18F-1 or 18F-2 were pre-incubated for

60 min or 180 min, respectively. Dissociation was started by the addition of 30 µM unlabeled 1

or 2 in the incubation buffer. Samples were taken at 0.5, 1, 3, 5, 10, 20, 30, 45, and 60 min after

incubation with 18F-1 or18F-2, and additional samples were taken at 90, 120, and 180 min of

incubation with 18F-1. Receptor-bound ligand was separated from free ligand by filtration as

described above. Filter-bound radioactivity was measured by gamma-counting (Wallac Wizard

1480, PerkinElmer, Rodgau, Germany) and specific binding at various times calculated.

7

By non-linear regression analyses (GraphPad Prism 3.0, GraphPad Software Inc., La

Jolla, CA, USA), the observed rate constant kobs and the dissociation rate constant koff were

calculated from the association and dissociation experiments, respectively. The association rate

constant kon was calculated according to kon = (kobs-koff)/[radioligand], and the KD value

according to KD = koff/kon. Association half-time (Ass. t1/2) and dissociation half-time (Diss. t1/2)

were also calculated.

Results

The results from kinetic studies of 18F-1 and 18F-2 are presented in Supplemental Figures

6 and 7. Supplemental Table 1 lists the calculated kinetic parameters for 18F-1 and 18F-2. The

dissociation rate for 18F-1 was extremely low and could not be reliably measured. As a result, the

dissociation constant KD could not be determined for 18F-1.

Supplemental Table 1: In vitro kinetic parameters of 18F-1 and 18F-2

Radioligand σ1 receptor type

kon (M-1min-1)

koff (min-1)

KD (nM)

Ass. t1/2 (min)

Diss. t1/2 (min)

18F-1 human 0.0318 n.d n.d 22 > 120 rat 0.0318 n.d n.d. 47 > 180

18F-2 human 3.46 x 108 0.0342 0.099 7.1 20.3 rat 5.83 x 108 0.1213 0.208 3.9 5.7

n.d. = non-determinable

8

Supplemental Figure 6. Association (A, C) and dissociation (B, D) rate measurements of 8F-1

in rat cortex homogenates (A, B) and cloned human sigma-1 receptor (C, D).

9

Supplemental Figure 7. Association (A, C) and dissociation (B, D) rate measurements of 18F-2

in rat cortex homogenates (A, B) and cloned human sigma-1 receptor (C, D).

Documents

Journal of Nuclear Medicine, published on February 23, 2017 ......2017/02/22 · good binding characteristics in rodents, pigs, and a preliminary human study (14,16-19). Here, we